https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/155 Fri, 24 Jul 2026 03:02:33 GMT 2026-07-24T03:02:33Z https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/212945 Title: Development of machine learning-based site amplification models for Japan from borehole recordings Authors: Nguyen, Le-Anh-Nhat; Lee, Yong-Gook; Park, Duhee; Tsai, Chi-Chin Abstract: A large number of site amplification models have been developed using regression and machine learning (ML) approaches. Although ML models generally outperform traditional methods in predicting site amplification, the effects of specific site and motion parameters on model accuracy remain insufficiently explored. Using a meta-dataset of earthquake recordings from Japan’s Kiban Kyoshin Network (KiK-net), we trained six ML-based site amplification models: random forest (RF), extreme gradient boosting (XGB), and deep neural network (DNN), together with their hybrid variants incorporating Bayesian optimization (BO). A sensitivity analysis using RF evaluated how combinations of input proxies influence predictive performance, leading to the identification of an optimal proxy configuration. Among the six models trained with this configuration, BO-DNN performed best at period T < 0.1 s, whereas BO-XGB showed superior performance at T > 0.1 s. Shapley Additive exPlanations (SHAP) was used to rank proxy importance, identifying the peak frequency of the horizontal-to-vertical spectral-ratio curve (fp), the time-averaged shear-wave velocity up to 30 m (Vs30), borehole depth (BD), and borehole spectral acceleration averaged over 0.1–0.3 s (SS) as the most influential proxies. The proposed models demonstrate superior performance compared with two previously published models that were also developed using KiK-net data. Tue, 01 Dec 2026 00:00:00 GMT https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/212945 2026-12-01T00:00:00Z https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/218034 Title: S-scheme mediated charge transfer in titania-tungsten trioxide heterojunctions for photocatalytic mineralization of gaseous formaldehyde Authors: Lee, Chul-Seung; Maitlo, Hubdar Ali; Kim, Won-Ki; Lim, Dae-Hwan; Kim, Ki-Hyun Abstract: The prevalence of extended indoor activities has raised concerns about health risks associated with exposure to volatile organic compounds (VOCs). In response, WO3-coupled TiO2 S-scheme heterojunction photocatalysts are synthesized with varying nominal WO3 molar contents (1, 3, or 5 mol%). To clarify charge transfer mechanisms in TW-x heterojunctions during the photocatalytic degradation of VOCs like formaldehyde (FA), the interfacial electronic structure and charge-separation behavior of TiO2/WO3 composites are comprehensively assessed. To map these electronic properties, a suite of complementary techniques is employed, including ultraviolet photoelectron spectroscopy, Kelvin probe force microscopy, and Mott-Schottky analysis. The combined results provide converging evidence consistent with an S-scheme charge-transfer pathway. Among the composites, TW-3 exhibits optimal performance, achieving complete removal of 100 ppm FA (50% RH and a flow rate of 100 mL min−1) with a dynamic-clean air delivery rate of 600 L g−1 h−1 and an apparent quantum yield of 3.08%. In-situ DRIFTS analysis reveals a reaction pathway consistent with FA mineralization through key formate intermediates. This process ultimately leads to oxidation into CO2 and H2O, driven by highly reactive oxygen species generated during photocatalysis. Overall, this study not only presents a strategy for building advanced heterojunction materials but also provides a strengthened mechanistic framework for interpreting their charge-transfer behavior in indoor air purification. Thu, 01 Oct 2026 00:00:00 GMT https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/218034 2026-10-01T00:00:00Z https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/217780 Title: Physics-guided residual learning framework for aftershock time history prediction using mainshock acceleration data Authors: Chen, Mengdie; Mangalathu, Sujith; Jeon, Jong-Su Abstract: Accurate aftershock acceleration time history prediction is essential for post-earthquake structural evaluation, particularly under data-scarce conditions. A physics-guided residual learning framework is developed to generate aftershock acceleration time histories from mainshock acceleration time histories and static site parameters. The proposed approach begins with a pseudo-aftershock acceleration time history constructed by applying exponential attenuation to the mainshock, which serves as a physically inspired prior for residual learning. Multiscale convolutional encoders extract temporal features, whereas static variables such as magnitude and site conditions are fused through gated embedding. A residual correction module guided by cross-attention refines the pseudo-aftershock to match the observed aftershock responses. A hybrid loss function ensures consistency in both time and frequency domains. Despite the limited number of training samples, the proposed residual learning model achieved consistently low prediction errors and showed promising predictive performance under a limited-event setting based on 140 paired sequences from 10 seismic events, including 20 test sequences. The architecture further enabled interpretable forecasting, revealing the manner in which mainshock dynamics and static attributes jointly influenced residual adjustments. On an independent test set, the model achieved a mean absolute error of 0.00553 g (where g denotes the gravitational acceleration), a root mean square error of 0.01304 g, and a coefficient of determination of 0.612. These findings highlight the feasibility of residual-based physically guided forecasting for realistic structural demand assessments. Thu, 01 Oct 2026 00:00:00 GMT https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/217780 2026-10-01T00:00:00Z https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/218679 Title: From charge equilibrium to catalytic performance: Engineering the Fe2O3-TiO2 interface for optimal S-scheme VOC abatement Authors: Lim, Dae-Hwan; Maitlo, Hubdar Ali; Boukhvalov, Danil W.; Kim, Ki-Hyun Abstract: In this research, Fe2O3-TiO2 S-scheme heterojunction photocatalysts (x-FeT, where x is the molar % of Fe2O3) have been developed and employed for efficient abatement of gaseous toluene. Among the series, 0.25-FeT emerges as the optimal system, exhibiting superior photophysical properties: prolonged carrier lifetime (0.78 ns), narrowed bandgap (3.07 eV), high photocurrent density (27.4 μA cm−2), and low charge-transfer resistance (9.49 Ω). The composite achieves 76.1% toluene degradation with an apparent quantum yield of 3.39 × 10-2 % under optimal conditions (1 ppm toluene, dry air, 100 mL min−1, and 352 nm UV). Key to its performance is the deliberate engineering of the interfacial energetics: UPS and in situ XPS analysis reveal a strong internal electric field (IEF) and band bending characteristic of a robust S-scheme. In situ EPR confirms that the IEF promotes the selective recombination while preserving high-energy electrons in Fe2O3 (−0.34 V vs. NHE) and holes in TiO2 (+3.08 V vs. NHE), enabling prolific generation of •O2- and •OH radicals. In-situ DRIFTS, GC-MS, and DFT simulations elucidate the degradation pathway: hydrogen abstraction forms a benzyl radical, followed by oxidation, culminating in ring-opening and mineralization to CO2 and H2O. Precise control of the Fe2O3:TiO2 ratio tunes the Fermi-level alignment and IEF strength, thereby optimizing the S-scheme charge dynamics for efficient VOC oxidation. The 0.25-FeT heterojunction is thus recommended as a highly effective and rationally designed photocatalytic medium for the abatement of recalcitrant VOCs in air. Thu, 01 Oct 2026 00:00:00 GMT https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/218679 2026-10-01T00:00:00Z